Pain is one of the most common symptoms of pancreatic cancer (P.C.). In 1989 MSKCC became one of two institutions in an NCI sponsored Pancreas Pain Network. The RFA celled for the study of P.C. pain leading to the initiation of prospective random assignment trials comparing currently used pain relief methods. During the first two years of the grant, the MSKCC team assessed the incidence and prevalence of pain among newly diagnosed P.C. patients (pts). Using validated instruments including the MPAC, FLIC, WAC, STAI I and II, BECK and ISEL, 133 patients (89 prior to first surgery, 44 prior to first chemotherapy) were studied. At diagnosis, 27% of pts had no pain; for the 73% with it, diffuse poorly localized abdominal pain was the most common complaint. Although pain was common at study entrance, an analysis of pain intensity using the Tursky score demonstrated most pts had minimal to mild discomfort; only 26% of all pts had moderate, strong or severe pain (mod-severe) (with only 1% having severe pain). Similar results were found using a linear analogue test (VASPI). For those having it, pain relief was good to excellent with a median VASPR score of 81. While only a minority had mod-severe pain, pts with a Tursky >5 or VASPI >30 had a significantly lower WAC score. While depression is said to be common at diagnosis in pts with advanced P.C. this was not confirmed in the MSKCC study (median BECK score 11; only 26% of pts had scores of >15). Pain was an ominous prognostic indicator for preoperative pts (those with no pain had a 34% resection rate, while only 20% of those with any pain (Tursky >1) were resectable (p=0.08). While 70-75% of pts had, at study entrance, not only mild pain, we identified a substantial minority (25-30%) with moderate to severe pain (Tursky >5). Furthermore, in longitudinal follow-up, pts were studied for up to 14 months. 50% of those who presented with moderate to severe pain had persistent pain at this level despite standard oral analgesic intervention. Equally important, an additional 19% of pts who presented with minimal pain progressed to develop mod-severe pain. Thus, pts with poor control of pain can be identified for more intense therapy. MSKCC and Johns Hopkins University Hospital, (the other member of the pancreas pain network), have developed a prospective random assignment trial comparing two commonly used measures to relieve P.C. pain: radiation therapy versus a celiac block. Patients with poor control of pain (unacceptable Tursky plus VASPI) will first receive one week of a standardized intense oral pain regimen. Patients without adequate pain control will then be randomly assigned to receive either RT or a nerve block. The efficacy of the two modalities in relieving pain and psychosocial distress will be compared as will the effects on quality of life, the financial cost difference between the two treatments, and survival.